CN1200404C - Relative pulse position of code-excited linear predict voice coding - Google Patents
Relative pulse position of code-excited linear predict voice coding Download PDFInfo
- Publication number
- CN1200404C CN1200404C CNB011245921A CN01124592A CN1200404C CN 1200404 C CN1200404 C CN 1200404C CN B011245921 A CNB011245921 A CN B011245921A CN 01124592 A CN01124592 A CN 01124592A CN 1200404 C CN1200404 C CN 1200404C
- Authority
- CN
- China
- Prior art keywords
- signal
- pulse
- track
- signal pulse
- input
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 claims abstract description 16
- 230000005284 excitation Effects 0.000 claims abstract description 7
- 238000001914 filtration Methods 0.000 claims description 15
- 230000005540 biological transmission Effects 0.000 claims description 5
- 238000012546 transfer Methods 0.000 claims description 4
- 238000004891 communication Methods 0.000 description 18
- 230000015572 biosynthetic process Effects 0.000 description 14
- 238000003786 synthesis reaction Methods 0.000 description 12
- 238000010586 diagram Methods 0.000 description 8
- 230000006835 compression Effects 0.000 description 6
- 238000007906 compression Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 6
- 230000000737 periodic effect Effects 0.000 description 4
- 238000013507 mapping Methods 0.000 description 3
- 239000000203 mixture Substances 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000009432 framing Methods 0.000 description 1
- 230000000266 injurious effect Effects 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 230000037361 pathway Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001172 regenerating effect Effects 0.000 description 1
- 230000008929 regeneration Effects 0.000 description 1
- 238000011069 regeneration method Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
- G10L19/04—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using predictive techniques
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
- G10L19/04—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using predictive techniques
- G10L19/08—Determination or coding of the excitation function; Determination or coding of the long-term prediction parameters
- G10L19/10—Determination or coding of the excitation function; Determination or coding of the long-term prediction parameters the excitation function being a multipulse excitation
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
- G10L19/04—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using predictive techniques
- G10L19/08—Determination or coding of the excitation function; Determination or coding of the long-term prediction parameters
- G10L19/12—Determination or coding of the excitation function; Determination or coding of the long-term prediction parameters the excitation function being a code excitation, e.g. in code excited linear prediction [CELP] vocoders
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
An apparatus and method for vocoding an input signal comprising a linear predictive filter for generating a filtered signal with a first signal pulse and a second signal pulse in response to receiving the input signal and a processor having a lookup table with a plurality of track positions. The first signal pulse is associated with a first track position and the second signal pulse is associated with a second track position relative to the first signal pulse resulting in a plurality of excitation parameters.
Description
Technical field
The present invention relates to compress speech, be specifically related to Code Excited Linear Prediction (CELP) voice coding.
Background technology
" the A History of Engineering and Science in theBell System " the 99th that published in 1984, the situation (nineteen twenty-eight) that the 101-102 page or leaf has been described the invention of vocoder are in the 114th page of situation (1967) of having introduced linear forecast coding technology (LPC) of the document.The compressible speech signal of speech coders/decoders (vocoder, automatic speech compositor) is so that reduce transmission bandwidth required in the communication channel.Call out required transmission bandwidth by reducing each, in same communication channel, might increase the quantity of calling.In the previous speech coding technology, in all linear predictive codings in this way (LPC) technology, use a wave filter and remove signal redundancy, thus compressed voice signal.The LPC wave filter can reappear a kind of spectrum envelope, to attempt to imitate people's voice.In addition, the LPC wave filter is subjected to the excitation of the quasi periodic input that nasal sound and vowel aspect receive, and is received as the noise like input for the sound of non-voice.
One class vocoder is arranged, and known is Code Excited Linear Prediction (CELP) vocoder.The CELP voice coding mainly is a kind of speech data compress technique, and it can be comparable with the voice quality of other speech coding technologies on the 32kbps code check in the voice quality that can reach on the 4-8kbps code check.The previous LPC technology that the CELP vocoder compares has the improvement of two aspects.The first, the CELP vocoder is used a tone fallout predictor and is extracted tone information, attempts catching more voice details.The second, the CELP vocoder encourages the LPC wave filter with the noise-like signal of deriving in the residue signal that produces from the actual speech waveform.
Comprise three parts in the CELP vocoder: 1) short time predictive filter; 2) long-time predictive filter is also referred to as tone fallout predictor or self-adapting code book; And 3) fixed code book.The realization of compression is by the bit to each part assignment some, and they are less than the original used bit number of voice signal of expression.First's application linear prediction removes the short time redundancy in the voice signal.Generation is from the error of short time fallout predictor or be the echo signal that residue signal becomes long-time fallout predictor.
The speech language has a kind of character of quasi periodic, and long-time fallout predictor extracts a pitch period from residue signal, and gets rid of predictable information in the past one-period.After long-time and short time predictive filter, resulting residue signal almost is noise-like signal entirely.Use a kind of analysis, integrated approach, seek a kind of optimum matching, from its vector storehouse, replace this noise like residue signal with an input item by a fixed codebook search.Represent the code word of optimum matching just to replace this noise like residue signal and transmit.In algebraically CELP (ACELP) vocoder, fixed code book is made up of a spot of non-zero pulses, and is represented with the position of pulse and symbol (for example+1 or-1).
In a kind of typical scheme, the CELP vocoder is with the input speech signal piecemeal or divide framing, each frame is made the LPC coefficient update of a short time fallout predictor.Then, the LPC residue signal is divided into subframe, is used for long-time fallout predictor and fixed codebook search.For example, for the short time fallout predictor, the input voice can be blocked into the frame of 160 samples.Then, the frame of formation can be divided into three subframes of 53 samples, 53 samples and 54 samples.So each subframe is handled by long-time fallout predictor and fixed codebook search.
Referring to Fig. 1, it shows the example of voice signal 100 single subframes.Voice signal 100 is made up of the voice and the non-speech audio of different tones.This voice signal 100 is received by a CELP vocoder with a kind of LPC wave filter.The CELP vocoder first step is the short time redundancy of removing in the voice signal.Remove the signal that obtains after the short time redundancy and be the remaining voice signal 200 among Fig. 2.
The LPC wave filter can not be removed whole redundant informations, and remaining quasi periodic peak point and trench are called tone pulses in the voice signal 200 of filtering.Then, on the short time predictive filter, add voice signal 200, obtain among Fig. 3 signal 300 through short time filtering.Long-time predictive filter is removed the tone pulses of quasi periodic from the remaining voice signal 300 of Fig. 3, forming among Fig. 4 almost is the signal 400 of noise like entirely, and it becomes the echo signal of fixed codebook search.Fig. 4 is the curve of the fixed code book echo signal of one 160 sample frame, and it is divided into three subframes 354,356,358.Then, this code word value is transmitted in communication network.
Among Fig. 5, show the look-up table 470 that pulse position in the subframe is shone upon.Each pulse in the subframe is restrained be positioned at one of 16 possible positions 402 of look-up table individual on.Owing in every track 404 16 possible positions 402 are arranged, so only need 4 bits can identify each pulse position.Each pulse mapping betides on each track 404.So two tracks 406,408 can be made mapping to the pulse position of two next signal pulses of subframe.
In the current example, the subframe 354 on Fig. 4 has only 53 samples in excitation, makes position 0-52 be only active position.Because track the 406, the 408th among Fig. 5, divided, so exist the position of the length that surpasses original excitation in every track.Position 56 in the track 1 and 60 and track 2 in position 57 and 61 be invalid and usefulness not.The position of two pulses 310,312 is corresponding to sample 13 and sample 17 among Fig. 4.By means of the table 470 of application drawing 5, can determine that sample 13 is arranged on the 3rd position 410 of article one track 406.Second pulse is positioned on the 4th position 412 of second track 408 in sample 17.So each pulse can be represented with 4 bits and be transmitted respectively.Owing to have only two tracks in the code book, can leave out of account to other pulse 314,316,318,320 and 322 in the subframe among Fig. 4 354.
Summary of the invention
Pulse position is lived by the absolute pulse position constraint in the track.Its shortcoming is, the CELP vocoder is placed into pulse on the adjacent position in the track easily.When being placed into pulse on the adjacent position in the track, will to spoken sounds begin encode rather than intonation carried out the more coding of balance.In addition, when the bit rate that is used for vocoder reduces and seldom pulse can use the time, the poor efficiency that track is put in pulse makes voice quality be subjected to injurious effects.Need a kind of method, it can reduce pulse and be placed into incidence in the adjacent track position.
By position with respect to a signal pulse in first track, a signal pulse is placed in one second track, can eliminate the poor efficiency of absolute orbit location layout.During signal pulse of coding, the N+1 signal pulse in the N+1 track is implemented relative positioning, can make decoded signal that the signal quality that increases is arranged.The realization of the signal quality that increases is by means of more accurately pulse being placed in the track, and the incidence that reduces signal pulse position placed adjacent in track.
Description of drawings
With reference to several accompanying drawings and from detailed description of the present invention, will understand above-mentioned purpose of the present invention and beneficial characteristics apparently, in each accompanying drawing:
Fig. 1 example goes out a frame of voice signal;
Fig. 2 example goes out a speech frame through short cycle filtering;
Fig. 3 example goes out a speech frame through self-adapting code book filtering;
Fig. 4 example goes out a kind of known method, and the speech frame of 160 compositions of sample is divided into three subframes;
Fig. 5 is the diagram of a known CELP vocoder code book look-up table, and signal pulse constrains on one of 16 possible pulse positions;
Fig. 6 is the diagram of a CELP vocoder code book, has the pulse position according to the relative restraint of one embodiment of the invention in the code book;
Fig. 7 is the diagram of a communication system, and it has emitter and the receiving trap that carries out the CELP encoding and decoding speech according to one embodiment of the invention;
Fig. 8 is a diagram with emitter of CELP vocoder, and this CELP vocoder is encoded to voice signal according to one embodiment of the present of invention;
Fig. 9 is the receiving trap diagram that has according to the CELP vocoder of one embodiment of the invention;
Figure 10 is according to one embodiment of the invention, voice signal is carried out a process flow diagram of the method for voice coding.
Embodiment
Among Fig. 6, show the code book table of one two track, it has the pulse position of relative restraint.Comprise two pulse position tracks 502,504 (claiming that usually they are track) in the table 500, can identify 16 possible signal pulse positions 506 on each track.502 and track 2 by track 1 504 in the 13rd locational 508, fixed code book input item zero is possible effective impulse position there.On the pulse meter position in the code book 512 of the 510 and the 15th position of the 14th position, in two tracks, all do not use.In addition, in first track first possible pulse position be constrained to be in can by on 4 pulse positions that divide exactly (also promptly 0,4,8 ..., on 52).Second pulse position in second track is that the index location 506 with first signal pulse in first track is as the criterion.
Not that signal pulse is coded in the adjacent track position, but produce a relative positioning of secondary signal pulse.By means of making the adjacent signals pulse of encoding in the track seldom, each signal pulse can reappear sudden energy preferably, and this has just improved the voice quality by the signal of vocoder decoding.In the present embodiment, single signal pulse is coded in each of two tracks 502 and 504.By the secondary signal pulsion phase in second track is positioned for first signal pulse in first track, can accomplish in the raising of having improved quality of decoding intonation.Among another embodiment, comprise plural track in the code book table, the additional signal pulse in each track is as the criterion with the previous track position of previous signal pulse.
In this enforcement, the relative position of secondary signal pulse is for first signal pulse in first track in second track.Among another embodiment, the relative position of secondary signal pulse is for the first signal pulse sample position in second track.Again in another embodiment, the signal pulse position in second track can divide into groups with the order of non-order (just, 1 ,-1,7 ,-7,2 ,-2,6 ,-6,3 ,-3,5 ,-5,4 ,-4).
Forward on Fig. 7, it shows the communication system 600 with emitter 602 and receiving trap 604.Emitter 602 is linked together by communication path 606 with receiving trap 604.Communication path 606 can be selected a cable network (such as LAN (Local Area Network), wide area network, the Internet, ATM net or public telephone network) or a wireless network (such as Cellular Networks, microwave net or satellite network).To the major requirement of communication path 606 be can be between emitter 602 and receiving trap 604 transmission of digital data.
Each other signal input and output unit 608 and 610 is arranged on emitter 602 and the receiving trap 604.Unit 608 and 610 all is illustrated as telephone device, transmission of analogue signal to and fro between they and emitter 602 and the receiving trap 604.Signal I/O unit 608 is connected with emitter 602 by one two line communication path 612.Similarly, another signal I/O unit 610 is connected with receiving trap 604 through another two lines communication path 614.Among another embodiment, signal input unit is incorporated in (also being that they are to be produced on loudspeaker and the microphone that transmits and receives in the device) in the communicator that transmits and receives usefulness, perhaps intercoms mutually via wireless communications pathway (also being wireless phone).
Contain one in the emitter 602 and be connected to the simulating signal port 616 on the two line communication paths 612, a CELP vocoder 618 and a controller 620.Controller 620 is connected with simulating signal port 616, vocoder 618 and network interface 622.In addition, network interface 622 is connected with vocoder 618, controller 620 and communication path 606.
Similarly, contain another network interface 624, communication path 606 and another vocoder 628 that is connected on another controller 626 in the receiving trap 604.This another controller 626 is connected with another vocoder 628, another network interface 624 and another simulating signal port 630.In addition, this another simulating signal port 630 is connected on another two lines communication path 614.
On analog port 616, be received from the voice signal of signal input apparatus 608.Controller 620 provides control signal and timing signal for emitter 602, and makes analog port 616 that the signal that receives is transferred to vocoder 618 to carry out signal compression.A fixed code book that is used to compress received signal is arranged in the vocoder 618, and the data structure of code book is shown among Fig. 6.The data structure 500 of Fig. 6 is associated an interior pulse position of first signal pulse that filtering signal comes and first track.In addition, a secondary signal pulse is associated with one second pulse position, it be determined out with first track in the relativeness of first pulse position of first signal pulse.
By the relation of assignment second pulse position, avoid two signal pulse assignment position adjacent relations with respect to first pulse position.First signal pulse is encoded, and assignment makes the pulse position of secondary signal pulse in second track 504 encode with respect to first track 502 again with a pulse position in first track 502.The relative coding of second pulse position has obtained a compressed signal with bigger possibility, and promptly first pulse position is not adjacent with second pulse position.Then, the compressed signal that vocoder 618 provides among Fig. 7 is sent on the network interface 622.Network interface 622 makes compressed signal transfer on the receiving trap 604 via communication path 606.
Another network interface 624 that is positioned on the receiving trap 604 receives this compressed signal.Controller 626 in the receiving trap transfers on the vocoder 628 in the receiving trap compressed signal of reception.Vocoder 628 utilizes the 500 pairs of compressed signals of look-up table among Fig. 6 to decode.Among Fig. 7, the look-up table 500 in vocoder 628 application drawings 6 bears a simulating signal again from the compressed signal that receives.This look-up table 500 recovers the base value that fixed code book has, and carries out filtering with long-time and short time fallout predictor then.The simulating signal that obtains is sent on the signal I/O unit 610 of receiver end by the receiving trap simulating signal port 630 among Fig. 7.
Forward on Fig. 8, it shows the signal Processing of the analog voice signal of being implemented by emitter 602.An input end that receives simulating signal is arranged on the pretreater 710, and its output is fed on LP wave filter 714 and the signal mixer 712.Signal mixer 712 mixes the signal of pretreater 710 and synthesis filter 716.The output of signal mixer 712 is fed on the perceptual weighting processor 718.Synthesis filter 716 is connected with tone analysis device 722 with LP analysis filter 714, signal mixer 712, another signal mixer 720, self-adapting code book 732.Tone analysis device 722 is connected with parametric encoder 724 with perceptual weighting processor 718, fixed codebook search 734, self-adapting code book 732, synthesis filter 716, another signal mixer 720.Parametric encoder 724 is connected with transmitter 728, fixed codebook search 734, fixed code book 730, LP analysis filter 714 and tone analysis device 722.
On pretreater 710, receive simulating signal from simulating signal I/O unit 608 among Fig. 7.Among Fig. 8, pretreater 710 is handled this simulating signal, and regulates gain and other characteristics of signals.Then, the signal that provides of pretreater 710 is fed to LP analysis filter 714 and signal mixer 712 on both.The coefficient information that produces by LP analysis filter 714 (LPc ' info) be sent on synthesis filter 716, perceptual weighting processor 718 and the parametric encoder 724.Synthesis filter 716 receives from the LP coefficient information of LP analysis filter 714 with from the signal of another signal mixer 720.Roughly short time spectral shape in the synthesis filter 716 imitation voice produces a signal, and the output with pretreater 710 in signal mixer 712 mixes mutually.The signal that obtains from signal mixer 712 carries out filtering by perceptual weighting processor 718.Perceptual weighting processor 718 also receives the LP coefficient information of LP analysis filter 714.Perceptual weighting processor 718 is postfilters, there, by each frequency that contains high speech energy is amplified its signal spectrum, and those frequencies that comprise less speech energy is decayed, and can " shelter " coding distortion effectively.
The output of perceptual weighting processor 718 is sent on fixed codebook search 734 and the tone analysis device 722.The code word value that fixed codebook search 734 produces is sent on parametric encoder 724 and the fixed code book 730.Shown fixed codebook search 734 is separated with fixed code book 730, but can be included under the another kind of situation in the fixed code book 730, must not realize by separate mode.In addition, has visit mouth on the fixed codebook search 734 for the data structure of look-up table among Fig. 6 500, and its second pulse position is with respect to the judgement of first pulse position coded pulse signal information more accurately, and reduces in the code book incidence to the adjacent pulse coding.
Fixed code book 730 receives the code word value that is produced by fixed codebook search 734, bears a signal again.The signal that is produced mixes in signal mixer 720 with the signal of self-adapting code book 732.Then, the mixed signal of formation should be used for imitating short time spectral shape in the voice signal by synthesis filter 716, feeds back to self-adapting code book 732 again.
Among another embodiment of last plane system, the encoder in the vocoder partly coexists in all digital answering machines in this way of same device.In a kind of like this embodiment, communication path is a data bus, and it can make the signal of compression store in a storer and therefrom call.
Among Fig. 9, show the diagram that has the receiving trap 604 of a CELP vocoder according to one of the present invention embodiment.A network interface 616 that is connected on the receiver 802 is arranged in the receiving trap 604.A fixed code book 804 is connected on the receiver 802, and is connected on the gain coefficient " c " 812.Signal mixer 806 is connected with gain coefficient " c " 812 with synthesis filter 808 and gain coefficient " p " 811.Self-adapting code book 810 is connected on the gain coefficient " p " 811, and the output of received signal mixer 806.Synthesis filter 808 is connected to the output of signal mixer 806, and is connected on the rearmounted perceptual filter 814.Rearmounted perceptual filter 814 is connected with synthesis filter 808, and is connected on another analog port 630.
Receiving trap 604 receives the signal of compression on network interface 616.Data in the compressed signal that receives on 802 pairs of network interfaces 616 of receiver are unpacked.Comprise fixed code book index, fixed codebook gain, self-adapting code book index, self-adapting code book gain and the index that the LP coefficient is used in the data.Comprise the such look-up table 500 of data structure among a Fig. 6 in the fixed code book 804.Among Fig. 9, the signal that fixed code book 804 produces mixes mutually with the signal of self-adapting code book 810 through gain coefficient 812 in signal mixer 806.Then, the mixed signal that signal mixer 806 provides is received by synthesis filter 808, and feeds back to self-adapting code book 810.Synthesis filter 808 is used these mixed signals voice signal of regenerating.Voice signal obtains adjusting the voice signal of regeneration by rearmounted perceptual filter 814.Then, by analog port 630 voice signal is sent on the receiver, there is a similar code book there.
Forward Figure 10 to, it shows the process flow diagram of a voice coding method, wherein uses a kind of look-up table or code book, and the pulse position in the N+1 track is to be as the criterion with previous pulse position.In step 902, receive an input signal (for example, analog voice signal) by the receiving trap among Fig. 7 604.In the step 903 of Figure 10, input signal is divided into each signal frame, thereby can handle the signal section of separation.In the step 904 of Figure 10, handle each signal frame, obtain the input signal of a filtering, be called residue signal by the LP analysis filter among Fig. 8 714.
On the step 906 among Figure 10, the residue signal of filtering is further carried out filtering, and have long-time signal redundancy in the input signal of filtering of signal pulse by 732 pairs of the self-adapting code books among Fig. 8 and make translating into and removed by a long-time wave filter.On the step 908 among Figure 10, identify the position of first signal pulse in first track by the fixed code book index.Comprise a look-up table 500 among Fig. 6 in the fixed code book 730 among Fig. 8, and comprise in second track second pulse position for the relative mapping of first pulse position in first track.In the step 909, determine of the skew of second pulse position, and make second pulse that more accurate localization be arranged with respect to first pulse position.Look-up table 500 should be used for producing a binary data pattern by the fixed code book among Fig. 8 730, and it can represent remaining pulse signal in the input signal.Then, in the step 910 of Figure 10, the binary data pattern is encoded into a signal that contains the pulse position index.So, in step 912, transfer out this encoded signals through communication path.
Current state of the art can combine general purpose digital signal processor and other electronic unit, to make the CELP vocoder by software group structure.So, can comprise the software code word in a kind of computer-readable signal bearing media, to realize a kind of vocoder, it has additional constraint in order to the pulse position in the restriction code book.
Though, a kind of specific embodiment of reference at length shows bright and has described the present invention, but the skilled person in the present technique field understands, can make the various changes on formal and the details to this, they depart from not open the spirit and scope of the present invention, therefore, the usefulness of following claims is intended to, and all this kind changes all are in the claim scope.
Claims (13)
1. one kind is carried out the method for voice coding to input signal, comprises step:
Input signal is carried out filtering, obtain having the filtering signal of one first signal pulse and a secondary signal pulse;
First signal pulse is encoded, one first interior pulse position of first signal pulse and encoded data structure first track is associated; And
With respect to first pulse position, in this encoded data structure second track to non-conterminous second pulse position of this secondary signal pulse assignment.
2. the process of claim 1 wherein, comprise step in the filter step: handle this signal with a linear prediction filter.
3. the process of claim 1 wherein and become the step of a plurality of signal frames before filter step, to provide division of signal.
4. the method for claim 3, wherein, the step of division signals frame begins with the step that receives a simulating signal.
5. the method for claim 3, wherein, the step of division signals frame begins with the step that receives a digital signal.
6. the process of claim 1 wherein that the step of assignment pulse position comprises step: identify the offset of secondary signal pulse for first signal pulse.
7. the method for claim 6, wherein, the step of home position skew comprises step: calculate the skew of the first signal pulse position to a secondary signal pulse position.
8. equipment that is used for input signal is carried out voice coding comprises:
A linear prediction filter, it plays response to receiving inputted signal, produces the filtering signal that has one first signal pulse and a secondary signal pulse at least;
A processor, a look-up table that comprises a plurality of track position is arranged, wherein, in a plurality of track position of first to one first track position of the first signal pulse assignment, and in a plurality of pulse positions of second portion to second track position that is as the criterion with first track position of first signal pulse of secondary signal pulse assignment, obtain a plurality of excitation parameters thus; And
A transmitter, it transfers out this a plurality of excitation parameters in response to receive these a plurality of excitation parameters from processor in a transmission signals.
9. the equipment of claim 8 also comprises an input port, and it has one input signal is divided into the memory buffer unit of each input signal frame, to receive in response to the input port on the input port.
10. the equipment of claim 8 wherein, is determined in the signal of filtering the secondary signal pulse to an offset of first signal pulse by processor.
11. the equipment of claim 8 wherein, is determined the skew of secondary signal pulse to first track position by processor.
12. the equipment of claim 8, wherein, input signal is an analog input signal.
13. the equipment of claim 8, wherein, input signal is a digital signal.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US09/633,465 US6728669B1 (en) | 2000-08-07 | 2000-08-07 | Relative pulse position in celp vocoding |
| US09/633,465 | 2000-08-07 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1337671A CN1337671A (en) | 2002-02-27 |
| CN1200404C true CN1200404C (en) | 2005-05-04 |
Family
ID=24539732
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB011245921A Expired - Fee Related CN1200404C (en) | 2000-08-07 | 2001-08-06 | Relative pulse position of code-excited linear predict voice coding |
Country Status (9)
| Country | Link |
|---|---|
| US (1) | US6728669B1 (en) |
| EP (1) | EP1184842B1 (en) |
| JP (1) | JP5027966B2 (en) |
| KR (1) | KR20020012509A (en) |
| CN (1) | CN1200404C (en) |
| BR (1) | BR0106825A (en) |
| CA (1) | CA2350353A1 (en) |
| DE (1) | DE60101827T2 (en) |
| TW (1) | TW521265B (en) |
Families Citing this family (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6980948B2 (en) * | 2000-09-15 | 2005-12-27 | Mindspeed Technologies, Inc. | System of dynamic pulse position tracks for pulse-like excitation in speech coding |
| US6847929B2 (en) * | 2000-10-12 | 2005-01-25 | Texas Instruments Incorporated | Algebraic codebook system and method |
| KR100503414B1 (en) * | 2002-11-14 | 2005-07-22 | 한국전자통신연구원 | Focused searching method of fixed codebook, and apparatus thereof |
| US7742926B2 (en) | 2003-04-18 | 2010-06-22 | Realnetworks, Inc. | Digital audio signal compression method and apparatus |
| US20040208169A1 (en) * | 2003-04-18 | 2004-10-21 | Reznik Yuriy A. | Digital audio signal compression method and apparatus |
| DE602004007945T2 (en) * | 2003-09-29 | 2008-05-15 | Koninklijke Philips Electronics N.V. | CODING OF AUDIO SIGNALS |
| US8502706B2 (en) * | 2003-12-18 | 2013-08-06 | Intel Corporation | Bit allocation for encoding track information |
| KR100723400B1 (en) | 2004-05-12 | 2007-05-30 | 삼성전자주식회사 | Apparatus and method for encoding digital signal using plural look up table |
| SG123639A1 (en) * | 2004-12-31 | 2006-07-26 | St Microelectronics Asia | A system and method for supporting dual speech codecs |
| US7224295B2 (en) * | 2005-07-11 | 2007-05-29 | Mediatek Inc. | System and method for modulation and demodulation using code subset conversion |
| KR100900438B1 (en) * | 2006-04-25 | 2009-06-01 | 삼성전자주식회사 | Apparatus and method for voice packet recovery |
| US8688437B2 (en) | 2006-12-26 | 2014-04-01 | Huawei Technologies Co., Ltd. | Packet loss concealment for speech coding |
| CN101286321B (en) * | 2006-12-26 | 2013-01-09 | 华为技术有限公司 | Dual-pulse excited linear prediction for speech coding |
| CN102623012B (en) * | 2011-01-26 | 2014-08-20 | 华为技术有限公司 | Vector joint coding and decoding method, and codec |
| CN103098128B (en) * | 2011-06-15 | 2014-06-18 | 松下电器产业株式会社 | Pulse location search device, codebook search device, and methods therefor |
| US9546924B2 (en) | 2011-06-30 | 2017-01-17 | Telefonaktiebolaget Lm Ericsson (Publ) | Transform audio codec and methods for encoding and decoding a time segment of an audio signal |
| EP2763137B1 (en) * | 2011-09-28 | 2016-09-14 | LG Electronics Inc. | Voice signal encoding method and voice signal decoding method |
Family Cites Families (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4625286A (en) * | 1982-05-03 | 1986-11-25 | Texas Instruments Incorporated | Time encoding of LPC roots |
| NL8500843A (en) | 1985-03-22 | 1986-10-16 | Koninkl Philips Electronics Nv | MULTIPULS EXCITATION LINEAR-PREDICTIVE VOICE CODER. |
| US5754976A (en) | 1990-02-23 | 1998-05-19 | Universite De Sherbrooke | Algebraic codebook with signal-selected pulse amplitude/position combinations for fast coding of speech |
| CA2568984C (en) * | 1991-06-11 | 2007-07-10 | Qualcomm Incorporated | Variable rate vocoder |
| US5704003A (en) * | 1995-09-19 | 1997-12-30 | Lucent Technologies Inc. | RCELP coder |
| JP4063911B2 (en) * | 1996-02-21 | 2008-03-19 | 松下電器産業株式会社 | Speech encoding device |
| US5708757A (en) * | 1996-04-22 | 1998-01-13 | France Telecom | Method of determining parameters of a pitch synthesis filter in a speech coder, and speech coder implementing such method |
| US5924062A (en) * | 1997-07-01 | 1999-07-13 | Nokia Mobile Phones | ACLEP codec with modified autocorrelation matrix storage and search |
| WO1999010719A1 (en) * | 1997-08-29 | 1999-03-04 | The Regents Of The University Of California | Method and apparatus for hybrid coding of speech at 4kbps |
| US5963897A (en) | 1998-02-27 | 1999-10-05 | Lernout & Hauspie Speech Products N.V. | Apparatus and method for hybrid excited linear prediction speech encoding |
| US6067511A (en) * | 1998-07-13 | 2000-05-23 | Lockheed Martin Corp. | LPC speech synthesis using harmonic excitation generator with phase modulator for voiced speech |
| US6138092A (en) * | 1998-07-13 | 2000-10-24 | Lockheed Martin Corporation | CELP speech synthesizer with epoch-adaptive harmonic generator for pitch harmonics below voicing cutoff frequency |
| US6094629A (en) * | 1998-07-13 | 2000-07-25 | Lockheed Martin Corp. | Speech coding system and method including spectral quantizer |
| US6119082A (en) * | 1998-07-13 | 2000-09-12 | Lockheed Martin Corporation | Speech coding system and method including harmonic generator having an adaptive phase off-setter |
| JP4308345B2 (en) * | 1998-08-21 | 2009-08-05 | パナソニック株式会社 | Multi-mode speech encoding apparatus and decoding apparatus |
| US6240386B1 (en) * | 1998-08-24 | 2001-05-29 | Conexant Systems, Inc. | Speech codec employing noise classification for noise compensation |
| US6311154B1 (en) * | 1998-12-30 | 2001-10-30 | Nokia Mobile Phones Limited | Adaptive windows for analysis-by-synthesis CELP-type speech coding |
| US6539349B1 (en) * | 2000-02-15 | 2003-03-25 | Lucent Technologies Inc. | Constraining pulse positions in CELP vocoding |
-
2000
- 2000-08-07 US US09/633,465 patent/US6728669B1/en not_active Expired - Lifetime
-
2001
- 2001-06-13 CA CA002350353A patent/CA2350353A1/en not_active Abandoned
- 2001-07-02 EP EP01305719A patent/EP1184842B1/en not_active Expired - Lifetime
- 2001-07-02 DE DE60101827T patent/DE60101827T2/en not_active Expired - Lifetime
- 2001-07-31 BR BR0106825-3A patent/BR0106825A/en not_active Application Discontinuation
- 2001-08-02 TW TW090118919A patent/TW521265B/en active
- 2001-08-06 KR KR1020010047214A patent/KR20020012509A/en not_active Application Discontinuation
- 2001-08-06 CN CNB011245921A patent/CN1200404C/en not_active Expired - Fee Related
- 2001-08-07 JP JP2001239069A patent/JP5027966B2/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| DE60101827T2 (en) | 2004-12-02 |
| BR0106825A (en) | 2002-06-04 |
| US6728669B1 (en) | 2004-04-27 |
| EP1184842A3 (en) | 2002-05-15 |
| JP2002108400A (en) | 2002-04-10 |
| CA2350353A1 (en) | 2002-02-07 |
| EP1184842B1 (en) | 2004-01-21 |
| DE60101827D1 (en) | 2004-02-26 |
| CN1337671A (en) | 2002-02-27 |
| EP1184842A2 (en) | 2002-03-06 |
| KR20020012509A (en) | 2002-02-16 |
| TW521265B (en) | 2003-02-21 |
| JP5027966B2 (en) | 2012-09-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1200404C (en) | Relative pulse position of code-excited linear predict voice coding | |
| CN1223989C (en) | Frame erasure compensation method in variable rate speech coder | |
| CN1158647C (en) | Spectral magnetude quantization for a speech coder | |
| KR100487943B1 (en) | Speech coding | |
| CN101006495A (en) | Audio encoding apparatus, audio decoding apparatus, communication apparatus and audio encoding method | |
| CN1228867A (en) | Method and apparatus for improving voice quality of tandemed vocoders | |
| CN1470052A (en) | High frequency intensifier coding for bandwidth expansion speech coder and decoder | |
| JP2002328700A (en) | Hiding of frame erasure and method for the same | |
| WO2008007698A1 (en) | Lost frame compensating method, audio encoding apparatus and audio decoding apparatus | |
| US7634402B2 (en) | Apparatus for coding of variable bitrate wideband speech and audio signals, and a method thereof | |
| JP2003223189A (en) | Voice code converting method and apparatus | |
| EP1020848A2 (en) | Method for transmitting auxiliary information in a vocoder stream | |
| JP3063668B2 (en) | Voice encoding device and decoding device | |
| JP3964144B2 (en) | Method and apparatus for vocoding an input signal | |
| JP2001154699A (en) | Hiding for frame erasure and its method | |
| WO2003001172A1 (en) | Method and device for coding speech in analysis-by-synthesis speech coders | |
| US6385574B1 (en) | Reusing invalid pulse positions in CELP vocoding | |
| CN1347548A (en) | Speech synthesizer based on variable rate speech coding | |
| EP1527440A1 (en) | Speech communication unit and method for error mitigation of speech frames | |
| Soheili | Analysis by synthesis coding of speech signals at 8 kb/s and below | |
| JPH034300A (en) | Voice encoding and decoding system | |
| Keiser et al. | Parametric and Hybrid Coding | |
| GB2365297A (en) | Data modem compatible with speech codecs | |
| Campanella | VOICE PROCESSING TECHNIQUES | |
| JPH0683149B2 (en) | Speech band signal encoding / decoding device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20050504 Termination date: 20190806 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |